Modeling groundwater redox conditions at national scale through integration of sediment color and water chemistry in a machine learning framework.

Redox conditions play a crucial role in determining the fate of many contaminants in groundwater, impacting ecosystem services vital for both the aquatic environment and human water supply. Geospatial machine learning has previously successfully modelled large-scale redox conditions. This study is the first to consolidate the complementary information provided by sediment color and water chemistry to enhance our understanding of redox conditions in Denmark. In the first step, the depth to the first redox interface is modelled using sediment color from 27,042 boreholes. In the second step, the depth of the first redox interface is compared against water chemistry data at 22,198 wells to classify redox complexity. The absence of nitrate containing water below the first redox interface is referred to as continuous redox conditions. In contrast, discontinuous redox conditions are identified by the presence of nitrate below the first redox interface. Both models are built using 20 covariate maps, encompassing diverse hydrologically relevant information. The first redox interface is modelled with a mean error of 0.0 m and a root-mean-squared error of 8.0 m. The redox complexity model attains an accuracy of 69.8 %. Results indicate a mean depth to the first redox interface of 8.6 m and a standard deviation of 6.5 m. 60 % of Denmark is classified as discontinuous, indicating complex redox conditions, predominantly collocated in clay rich glacial landscapes. Both maps, i.e., first redox interface and redox complexity are largely driven by the water table and hydrogeology. The developed maps contribute to our understanding of subsurface redox processes, supporting national-scale land-use and water management.

National Assessment of Long-Term Groundwater Response to Pesticide Regulation.

Quantitative assessments of long-term, national-scale responses of groundwater quality to pesticide applications are essential to evaluate the effectiveness of pesticide regulations. Retardation time in the unsaturated zone (Ru) was estimated for selected herbicides (atrazine, simazine, and bentazon) and degradation products (desethylatrazine (DEA), desisopropylatrazine (DIA), desethyldesisopropylatrazine (DEIA), and BAM) using a multidecadal time series of groundwater solute chemistry (∼30 years) and herbicide sales (∼60 years). The sampling year was converted to recharge year using groundwater age. Then, Ru was estimated using a cross-correlation analysis of the sales and the frequencies of detection and exceedance of the drinking water standard (0.1 µg/L) of each selected compound. The results showed no retardation of the highly polar, thus mobile, parent compounds (i.e., bentazon), while Ru of the moderately polar compounds (i.e., simazine) was about a decade, and their degradation products showed even longer Ru. The temporal trends of the degradation products did not mirror those of the sale data, which were attributed to the various sale periods of the parent compounds, sorption of the parent compounds, and complex degradation pathways. The longer Ru in clayey/organic sediments than in sandy sediments further confirmed the role of soil-specific retardation as an important factor to consider in groundwater protection.

3D characterization of the subsurface redox architecture in complex geological settings.

Nitrogen (N) leaching caused by agricultural activities is one of the major threats to the aquatic ecosystems and public health. Moving from the agricultural soils through the subsurface and reemerging to the surface water, N undergoes various biogeochemical reactions along pathways in the subsurface, which occur heterogeneously in space and time. Thus to improve our understanding on the fate and distribution of N in the aquatic environment, detailed knowledge about the subsurface hydrogeological and biogeochemical conditions, especially the redox conditions, are essential. In this study, 3D information of the redox conditions termed the redox architecture was investigated in two Danish catchments with intensive agriculture underlain by glacial deposits. Towed transient electromagnetic (tTEM) resistivity was interpreted which reveals the subsurface geological structures at a few hectare scale. These geophysical data were integrated with sediment and water chemistry for the redox architecture interpretations. The top soils of both catchments are characterized as clay-till, but the tTEM showed that the subsurface hydrogeological structures are distinctively different. We identified three types of redox architectures in the studied catchments: 1) a planar redox architecture with a single redox interface; 2) a geological-window redox architecture with local complexity; and 3) a glaciotectonic-thrusted redox architecture with high complexity. The baseflow N load at the catchment outlets reflect the contributions of N via oxic pathways through the complex redox architectures of the subsurface. We conclude that in some landscapes, the redox architecture cannot be simplified as a single interface that roughly follows the terrain; hence, thorough investigations of the structural heterogeneity of the local redox architectures will be necessary to improve simulations of N evolution along pathways and quantifications of N attenuation under various mitigation scenarios.

Long-term nitrate response in shallow groundwater to agricultural N regulations in Denmark.

Over the last 30 years, Denmark has implemented a series of environmental action plans involving regulation of agricultural nitrogen (N) use and management in order to minimize the N pollution of the environment. The local effects of these action plan initiatives depend on various factors such as the efficiency of the implemented measures, and in particular, the hydrogeological structures and geochemical conditions of the subsurface that control the transport and fate of N. In this study, the effects of the Danish agricultural N regulations on shallow oxic groundwater are compared among five small agricultural catchments underlain by two types of lithology (sandy vs. loamy). Long-term spatially dense groundwater monitoring data is compared with monitoring data of nitrate in the root zone leaching and in stream. The results show clear improvements in the environmental state of shallow oxic groundwater in the first two decades since 1989 where the number of monitoring points with significant decreasing nitrate trends gradually increased for both soil types. Such improvements can be attributed to the effects of N mitigation measures implemented as a general regulation all over the country. However, deteriorations have been recorded in the last decade until 2016 where 26-35% of the monitoring points exhibited significant upward nitrate trends in both types of catchments. It is noteworthy that for sandy soils, the major part (93%) of the monitoring points showing significant upward trends in the period 2009-2016, also had concentrations above the groundwater standard of 50 mg/l nitrate in 2016. Altogether, the oxic groundwater in the sandy catchments was more responsive to N regulations than that in the loamy catchments. This might be due to efficient N regulation through statutory norms for the utilization of N in manure, increasing the N use efficiency in areas with a relatively high livestock density. Another reason is the nitrate vulnerability of the aquifers in sandy areas, with widespread oxic conditions from the top soil to the saturated zone. In the loamy catchments, nitrate may be reduced in near-surface localized reduced zones, and the reaction is often fast and the travel time from the root zone to the stream often relatively short. Therefore, stream nitrate concentrations were higher in the loamy catchments than in the sandy catchments. This is attributed to different hydrogeological pathways. Thus, in sandy catchments, deep groundwater is an important pathway, while in loamy catchments tile drains deliver nitrate directly to the streams. Our results indicate that N mitigation measures will help improve groundwater quality in sandy soils, while in loamy soils they will help to reduce surface water N loads. This implies that to achieve optimal environmental N reduction, agricultural N regulations should be strategically implemented according to farming characteristics and site-specific hydrogeological and geochemical conditions of the subsurface.

Groundwater nitrate response to sustainable nitrogen management.

Throughout the world, nitrogen (N) losses from intensive agricultural production may end up as undesirably high concentrations of nitrate in groundwater with a long-term impact on groundwater quality. This has human and environmental health consequences, due to the use of groundwater as a drinking water resource, and causes eutrophication of groundwater-dependent ecosystems such as wetlands, rivers and near-coastal areas. At national scale, the measured nitrate concentrations and trends in Danish oxic groundwater in the last 70 years correlate well with the annual agricultural N surpluses. We also show that the N use efficiency of agriculture is related to the groundwater nitrate concentrations. We demonstrate an inverted U-shape of annual nitrate concentrations as a function of economic growth from 1948 to 2014. Our analyses evidence a clear trend of a reversal at the beginning of the 1980s towards a more sustainable agricultural N management. This appears to be primarily driven by societal demand for groundwater protection linked to economic prosperity and an increased environmental awareness. However, the environmental and human health thresholds are still exceeded in many locations. Groundwater protection is of fundamental global importance, and this calls for further development of environmentally and economically sustainable N management in agriculture worldwide.

Pesticides in water supply wells in Zealand, Denmark: a statistical analysis.

Data from the Danish National Borehole Database are used to predict drinking water well vulnerability to contamination by pesticides, and to identify the dominant mechanisms leading to well pollution in Zealand, Denmark. The frequency of detection and concentrations of 4 herbicides and 3 herbicide metabolites are related to factors accounting for geology (thicknesses of sand, clay and chalk layers), geographical location (distance to surface water and distance to contaminated sites), redox conditions and well depth using logistic regression, the binomial test and Spearman correlation techniques. Results show that drinking water wells located in urban areas are more vulnerable to BAM and phenoxy acids contamination, while non-urban area wells are more subject to bentazone contamination. Parameters accounting for the hydraulic connection between the well and the surface (well depth and thickness of the clay confining layer) are often strongly related to well vulnerability. Results also show that wells close to surface water are more vulnerable to contamination, and that sandy layers provide better protection against the leaching of oxidizable pesticides than clay aquitards, because they are more likely to be aerobic. 4-CPP is observed more often at greater well depth, perhaps because of anaerobic dechlorination of dichlorprop. The field data are used to create a set of probabilistic models to predict well vulnerability to contamination by pesticides.

Trend reversal of nitrate in Danish groundwater--a reflection of agricultural practices and nitrogen surpluses since 1950.

This paper assesses the long-term development in the oxic groundwater nitrate concentration and nitrogen (N) loss due to intensive farming in Denmark. First, up to 20-year time-series from the national groundwater monitoring network enable a statistically systematic analysis of distribution, trends, and trend reversals in the groundwater nitrate concentration. Second, knowledge about the N surplus in Danish agriculture since 1950 is used as an indicator of the potential loss of N. Third, groundwater recharge CFC (chlorofluorocarbon) age determination allows linking of the first two data sets. The development in the nitrate concentration of oxic groundwater clearly mirrors the development in the national agricultural N surplus, and a corresponding trend reversal is found in groundwater. Regulation and technical improvements in the intensive farming in Denmark have succeeded in decreasing the N surplus by 40% since the mid 1980s, while at the same time maintaining crop yields and increasing the animal production of especially pigs. Trend analyses prove that the youngest (0-15 years old) oxic groundwater shows more pronounced significant downward nitrate trends (44%) than the oldest (25-50 years old) oxic groundwater (9%). This amounts to clear evidence of the effect of reduced nitrate leaching on groundwater nitrate concentrations in Denmark.